With the current I2C bus redundancy strategy, it is common to route two I2C buses to each field replaceable unit (FRU) included within a system. With this redundant topology, in the event that one I2C bus fails due to a short circuit occurring in one of the FRUs attached to one of the I2C buses (thus shorting the impacted I2C bus), access to the remaining FRUs (including the defective FRU itself provided the redundant I2C interface is operational) is maintained due to access being available via the redundant I2C bus.
However, one shortcoming of this design is the inability to determine which FRU has failed, as the only information that the field engineer knows is that one FRU on the I2C bus is shorting the I2C bus. Accordingly, the field engineer would need to remove each FRU on the impacted I2C bus, one at a time, and check the status of the impacted I2C bus during each removal to see if the impacted I2C bus was restored. As several I2C devices may be coupled to the impacted I2C bus, this could prove to be a time consuming and fallible procedure for the field engineer.